ADDED EXPLANATION:

Duplicate detection logic varies by survey type. For emergence count surveys, each unique combination of Site Name, Exit Identifier (if provided), Surveyor, and Species should appear in a single row—multiple rows from the same surveyor for the same combination are flagged as duplicates, whereas rows from different surveyors are averaged to produce the final count. Roost count surveys use different logic: counts are summed within each surveyor's report for a given site, section, and species, and if multiple surveyors report, their sums are averaged—so duplicates are not flagged in the same way. Beyond survey-type logic, duplicate errors can arise from re-uploading the same data in quick succession, inadvertently uploading the same survey under different file names, accidental copy/paste duplication within a file, or leftover data from older versions of the upload logic. When identifying duplicates within a file, Excel's Conditional Formatting (Home → Styles → Conditional Formatting → Highlight Cells → Duplicate Values) can highlight repeated values in a unique identifier column. For surveys lacking a single unique field, create a new column that concatenates key fields to form a composite identifier. When comparing across uploads, note the conflicting file's name and created date in the Partner Portal, download it, and review side-by-side with your current file to determine which version is more complete or accurate. In rare cases, the database may retain artifacts of previously deleted data that trigger duplicate errors on subsequent uploads—if you suspect this scenario after repeated removal and re-upload attempts, contact the NABat technical outreach team to investigate.